Overegging it
Q: Did you see the headline in Stuff that eggs actually make you kinder?
A: Yes.
Q: A bit ironic, isn’t it?
A: You mean because of how eggs are usually produced?
Q: Yes. Did they use free-range eggs, or battery farmed?
A: No eggs were harmed in the research.
Q: But it says “eggs” in the headline
A: Yes, it does.
Q: Was it another mouse lab study?
A: No, this was real people (university students) in a randomized experiment. The story says that, and even links to the paper.
Q: So it’s real?
A: I wouldn’t go that far.
Q: But experiment. And causal. And Science. Yes?
A: Very small experiment with marginally-significant results that could easily be due to chance.
Q: You’ve got a thing about psychologists, like Andrew Gelman does, haven’t you?
A: No, I’ve got a thing about over-promoted, under-powered research, like Andrew Gelman does.
Q: Ok, let’s calm down and get back to the study. What did they use instead of eggs?
A: 800 mg of tryptophan powder (or placebo) in orange juice.
Q: Is that a lot?
A: They say it’s the equivalent of three eggs.
Q: Why do you sound dubious?
A: Because other websites think there’s less than that. I haven’t found any definitive primary source, but there’s about 6g of protein in a 50g egg, and from Table 2 in this 1973 paper you can compute that egg protein is about 1.8% tryptophan, which works out as 108mg. Or in a 70g egg (‘Jumbo’ in NZ, ‘Large’ in Europe) about 150mg. It looks like you’d need more than five 70g eggs to get 800mg of tryptophan.
Q: How about other foods? Are eggs especially high in tryptophan?
A: Not especially. Chicken, or cheese, or oats, or chickpeas, for example, have more tryptophan per 100g.
Q: So if they didn’t use eggs in the experiment, and eggs aren’t particularly high in tryptophan, why are the headlines about eggs? Were they sponsored by the Egg Foundation?
A: No, nothing like that. And the university news post just talks about “the amino acid tryptophan, found in fish, soya, eggs and spinach.” It’s probably because the dose in the study was described in terms of eggs.
Thomas Lumley (@tslumley) is Professor of Biostatistics at the University of Auckland. His research interests include semiparametric models, survey sampling, statistical computing, foundations of statistics, and whatever methodological problems his medical collaborators come up with. He also blogs at Biased and Inefficient See all posts by Thomas Lumley »